Influence of the nature of the absorption band on the potential performance of high molar extinction coefficient ruthenium(II) polypyridinic complexes as dyes for sensitized solar cells

Article abstract of DOI:10.1021/ic1020862

When tested in solar cells, ruthenium polypyridinic dyes with extended π systems show an enhanced light-harvesting capacity that is not necessarily reflected by a high (collected electrons)/(absorbed photons) ratio. Provided that metal-to-ligand charge transfer bands, MLCT, are more effective, due to their directionality, than intraligand (IL) π-π* bands for the electron injection process in the solar cell, it seems important to explore and clarify the nature of the absorption bands present in these types of dyes. This article aims to elucidate if all the absorbed photons of these dyes are potentially useful in the generation of electric current. In other words, their potentiality as dyes must also be analyzed from the point of view of their contribution to the generation of excited states potentially useful for direct injection. Focusing on the assignment of the absorption bands and the nature of the emitting state, a systematic study for a series of ruthenium complexes with 4,4′-distyryl-2,2′-dipyridine (LH) and 4,4′-bis[p- (dimethylamino)-α-styryl]-2,2′-bipyridine (LNMe₂) "chromophoric" ligands was undertaken. The observed experimental results were complemented with TDDFT calculations to elucidate the nature of the absorption bands, and a theoretical model was proposed to predict the available energy that could be injected from a singlet or a triplet excited state. For the series studied, the results indicate that the percentage of MLCT character to the anchored ligand for the lower energy absorption band follows the order [Ru(deebpy)₂(LNMe₂)](PF₆)₂ > [Ru(deebpy)₂(LH)](PF₆)₂ > [Ru(deebpy)(LH)₂](PF₆)₂, where deebpy is 4,4′-bis(ethoxycarbonyl)-2,2′-bipyridine, predicting that, at least from this point of view, their efficiency as dyes should follow the same trend.

Full text of DOI:10.1021/ic1020862

ARTICLE
pubs.acs.org/IC
Influence of the Nature of the Absorption Band on the Potential
Performance of High Molar Extinction Coefficient Ruthenium(II)
Polypyridinic Complexes As Dyes for Sensitized Solar Cells
†
†
†
‡
,†
Francisco Gajardo, Mauricio Barrera, Ricardo Vargas, Irma Crivelli, and Barbara Loeb*
†
Facultad de Química, Pontificia Universidad Cat ꢀo lica de Chile, Casilla 306, Santiago, Chile
‡
Facultad de Ciencias, Universidad de Chile, Las Palmeras 3425, Santiago, Chile
S Supporting Information
b
ABSTRACT: When tested in solar cells, ruthenium polypyr-
idinic dyes with extended π systems show an enhanced light-
harvesting capacity that is not necessarily reflected by a high
(collected electrons)/(absorbed photons) ratio. Provided that
metal-to-ligand charge transfer bands, MLCT, are more effec-
tive, due to their directionality, than intraligand (IL) πꢀπ*
bands for the electron injection process in the solar cell, it seems
important to explore and clarify the nature of the absorption
bands present in these types of dyes. This article aims to
elucidate if all the absorbed photons of these dyes are potentially
useful in the generation of electric current. In other words, their potentiality as dyes must also be analyzed from the point of view of
their contribution to the generation of excited states potentially useful for direct injection. Focusing on the assignment of the
0
absorption bands and the nature of the emitting state, a systematic study for a series of ruthenium complexes with 4,4 -distyryl-2,
0
0
0
2
-dipyridine (LH) and 4,4 -bis[p-(dimethylamino)-R-styryl]-2,2 -bipyridine (LNMe ) “chromophoric” ligands was undertaken.
2
The observed experimental results were complemented with TDDFT calculations to elucidate the nature of the absorption bands,
and a theoretical model was proposed to predict the available energy that could be injected from a singlet or a triplet excited state.
For the series studied, the results indicate that the percentage of MLCT character to the anchored ligand for the lower energy
absorption band follows the order [Ru(deebpy) (LNMe )](PF ) > [Ru(deebpy) (LH)](PF ) > [Ru(deebpy)(LH) ](PF ) ,
2
2
6 2
2
6 2
2
6 2
0
0
where deebpy is 4,4 -bis(ethoxycarbonyl)-2,2 -bipyridine, predicting that, at least from this point of view, their efficiency as dyes
should follow the same trend.
’
INTRODUCTION
the point of view of the number of photons absorbed or by the
amplitude of the wavelength interval where absorption occurs.
On this basis, some ruthenium dyes have been reported present-
ing “chromophoric polypyridinic ligands”, e.g., bipyridinic li-
gands with an attached extended π-electronic system, Scheme 1.
On the other hand, donor groups such as methoxy or substituted
Ruthenium compounds containing monodentate, bidentate,
tridentate, and tetradentate pyridinic ligands have been intensely
used as dyes in nanocrystalline sensitized solar cells. These pyridinic
and polypyridinic ligands can be functionalized by incorporating
electron-withdrawing groups (e.g., carboxylate, phosphonate),
which at the same time play the role of an anchor group to the
TiO semiconductor in the cell. The ruthenium polypyridinic
5
amines have been incorporated as substituents on these “chro-
2a,b,6
7
1
mophoric ligands”,
allowing an increase in absorption.
2
Specifically, a series of ruthenium complexes with ligands of the
complexes present intense metal-to-ligand charge transfer
0
0
2
type 4,4 -distyryl-2,2 -dipyridine have been reported, presenting
an increased light-harvesting capacity and a relatively good yield
as dyes. In 2006 Gr €a tzel et al. described the requirements for
light-harvesting systems, emphasizing the directionality of the
charge redistribution produced by light absorption. Experimental
results have been reported for the tetrabutylammonium
(
MLCT) absorption bands in the visible region, which is one
of the factors directly responsible for optimization of the absorp-
3
tion processor LHE (light-harvesting efficiency) of the cell. Thus,
an increase in the molar extinction coefficient of this band in the
ruthenium dye chromophore is expected to increase the LHE of
the cell. This effect in turn could have a positive impact on the
IPCE (incident photon to collected electron quantum efficiency)
cell parameter and consequently on the photoaction spectra of
0
0
[
(
ruthenium (4-carboxylic acid-4 -carboxylate-2,2 -bipyridine)-
0
0
4,4 -di(2-(3,6-dimethoxyphenyl)ethenyl)-2,2 -bipyridine)(NCS) ],
2
4
N945H, and the fully deprotonated tetrabutylammonium
IPCE (λ) vs λ.
An increase of the capacity of the dye to absorb radiation appears
as one of the key factors to improve solar energy conversion to
electricity by a solar cell. This aspect can be undertaken either from
Received: October 14, 2010
Published: May 31, 2011
r 2011 American Chemical Society
5910
dx.doi.org/10.1021/ic1020862
|Inorg. Chem. 2011, 50, 5910–5924

Inorganic Chemistry
ARTICLE
Scheme 1
Generally, it is accepted that MLCT, due to its directionality,
should be effective for electron injection in the semiconductor
band, while IL bands should not. Published articles in the
literature related to ruthenium complexes with “chromopho-
ric polypyridinic ligands” are not conclusive on assigning the
absorption bands, fundamentally for the fact that the IL bands
1
1
shift to lower energies for donor substituents in the ligand.
In this case, the energy of the IL band may come close to the
energy level of the MLCT band and even produce a crossing
between these two levels, making difficult an assignment by a
2
þ
simple correlation with the spectrum of [Ru(bpy) ] -type
complexes.
3
1
2
Regarding another experimental variable, Filippo De Angelis
et al. published a manuscript about the influence of the sensitizer
adsorption mode on the open-circuit potential of dye-sensitized
solar cells. In order to understand the origin of such experimental
observations, they performed DFT calculations to obtain infor-
mation on the dipole moments in the geometry and orientation
corresponding to the free and adsorbed dye for the different
isomers of (Bu N) [Ru(dcbpyH) (NCS) ], N719, and [Ru-
4
2
2
2
(
2
dcbpyH )(tdbpy)(NCS) ], N621, sensitizers, with dcbpy =
2 2
0 0 0
,2 -bipyridyl-4,4 -dicarboxylato, dcbpyH = (4,4 -dicarboxylic
2
8b
0
0
0
acid-2,2 -bipyridine), and tdbpy = 4,4 -ditridecyl-2,2 -bipyridine.
Also reported was the relation of the magnitude and orientation of
dipolar fields with the different adsorption modes and with the
value of the open-circuit voltage.
Finally, the injection capacity of the dye is also an important
variable to be considered in a solar cell. This injection may occur
from a FranckꢀCondon “hot” excited state or, more probable,
from the thermally equilibrated “thexi” state. For a ruthenium
complex with polypyridinic ligands, the latter is probably the
lowest lying excited triplet state.
The present article aims to elucidate whether the absorbed
photons by this variety of dyes with “chromophoric ligands” are
potentially useful in the generation of electric current. Focusing
on the assignment of the absorption bands of this sort of
complexes, a systematic study using spectroscopic as well as
electrochemical tools was undertaken. Also examined was the
effect of the presence of donor substituents in the chromophoric
ligands on the relative position of the MLCT and the IL bands.
The observed experimental results were analyzed with theore-
13
tical calculations. Specifically, the nature of the electronic
transitions was examined by simulating their electronic spectra
with TDDFT methods and calculating the transition density
change within the molecule when an electronic transition occurs.
An electron density fragment analysis was performed, in order to
understand the electronic density distribution in the different
ligands, both for the first excited singlet and for the first excited
triplet states. The direct injection capacity of these states was
analyzed based on the electronic density on the ligand directly
anchored to the semiconductor surface.
0
0
0
[
ruthenium (4,4 -carboxylate-2,2 -bipyridine)(4,4 -di(2-(3-dim-
0
ethoxyphenyl)-ethenyl)-2,2 -bipyridine)(NCS) ], N945, sensi-
tizers, together with an INDO/DFT modeling of the N945
complex absorbed on TiO . In this report the authors conclude
that selective functionalization of a ruthenium complex allows
red-shift absorption spectra and an enhanced molar extinction
coefficient and, at the same time, creating significant direction-
ality in the excited state.
2
8a
2
To achieve this goal, a series of ruthenium complexes bearing
0
0
“chromophoric ligands” of the type 4,4 -distyryl-2,2 -dipyridine,
0
0
LH, and 4,4 -bis[p-(dimethylamino)-R-styryl]-2,2 -bipyridine,
LNMe , was synthesized. The ligands and the corresponding
2
In spite of the mentioned increased light-harvesting capacity
of complexes with these types of “chromophoric ligands”, the
estimated internal quantum efficiency (ratio of collected elec-
trons to absorbed photons) for these dyes is only 50%, which is
complexes are depicted in Scheme 1. A comparison of different
pairs of complexes allows analyzing the effect of different factors.
Specifically, by comparing complexes [Ru(bpy) (LH)](PF )
2
6 2
(1) and [Ru(bpy) (LNMe )](PF ) (2) or [Ru(deebpy) (LH)]-
2
2
6 2
2
9
lower than expected according to their high absorption. To shed
(PF ) (3) and [Ru(deebpy) (LNMe )](PF ) (4) the effect of
6 2 2 2 6 2
some light into the latter point, it is relevant to determine the
nature of the absorption bands, especially in the visible region.
the donor NMe group on the nature of the lowest energy band
was analyzed. On the other hand, by comparing complexes 3 and
2
10
5
911
dx.doi.org/10.1021/ic1020862 |Inorg. Chem. 2011, 50, 5910–5924

Inorganic Chemistry
ARTICLE
[
(
Ru(deebpy)(LH) ](PF ) (5) or 4 and [Ru(deebpy)(LNMe ) ]-
2 6 2 2 2
PF ) (6) the presence of the anchoring carboxylate groups on
6
2
the electronic nature of the complexes was analyzed. Also studied
were the spectroscopic properties of complexes 5 and 6 with an
increased number of chromophoric ligands attached to the
ruthenium center. As mentioned, the character of the absorption
band was elucidated by means of TDDFT-type calculations, and
the predominance of the MLCT absorption to the anchoring bpy
ligand as well as a high electronic concentration in this ligand in
the excited S and T were related to a probable higher efficiency as
a solar cell dye.
Figure 1. Decomposition of the chromophoric ligand in terms of
orthogonal fragments.
were checked to be at an energy minimum by performing vibrational
frequencies calculations and verifying that only real frequencies were
present. Solvations effects were also included by means of the
COSMO model.
’
EXPERIMENTAL SECTION
Analytical Methods. UVꢀVis and Emission Spectroscopy. UVꢀvis
0
0
absorption measurements were made on a Shimadzu UV 3101PC spectro-
photometer. Photoluminescence (PL) spectra were recorded on a Perkin-
Elmer L55 spectrofluorophotometer.
Fragment Analysis. Because the 4,4 -distytyl-2,2 -dipyridine-type
ligands show a C symmetry they can be analyzed considering one-half
2
of it. Additionally, it is possible to define three kinds of orthogonal
fragments, F, V, and P, Figure 1. The fragment F contains the para-
substituted phenyl ring, fragment V is centered on the vinyl moiety, and
P corresponds to the pyridine ring.
Infrared Spectroscopy. Infrared (IR) spectra for the desired com-
pounds and reactants were recorded as KBr mulls in a Bruker Vector 22
FTIR spectrometer.
Elemental Analysis and Mass Spectrometry. Elemental analysis was
performed on a Fisons Instrument Analyzer, model EA1108/CHNS-O
with PC NCR system 3225. Mass spectra were recorded with an LCQ
Duo Ion Trap Mass Spectrometer (Thermo Finnigan, USA).
The same scheme was applied for the complexes under study: the C
2
symmetry enables one to decompose the whole structure in terms of five
orthogonal fragments; the first two are centered on the substituted
bipyridine and on the ruthenium atom. The remaining three fragments
are located on the chromophoric ligand and follow the decomposition
mentioned above. Through this approach it is possible to assign a
character to each molecular orbital in accordance to the predominant
fragment(s), leading to classifying them as metallic (M) or centered on
bipyridine (B), on the dimethyl ester of bipyridine (Bp), or on the
chromophoric ligands (C).
1
NMR Spectroscopy. H NMR spectra were recorded on a Bruker ACL
3 4
200 200 MHz spectrometer with tetramethylsilane, Si(CH ) , as the
reference or on a Bruker AVANCE 400 FT-NMR spectrometer.
Electrochemistry. Cyclic voltammetry was performed in nitrogen-
saturated 0.1 M tetrabutylammonium hexafluorophosphate (TBAPF
6
)
as supporting electrolyte in CH CN. A BAS model CV50w potentiostat
3
was used in a standard three-electrode arrangement with a Pt or glassy
platinum working electrode, a Pt gauze counter electrode, and a Ag/
AgCl reference electrode. All potentials are referred with respect to this
last electrode.
Transition Density Analysis. Molecular orbitals (MO) can be decom-
posed in terms of a linear combination of orthogonal fragments, each of
them related to a specific group of atoms with chemical meaning (ligand,
substituent, or metal)
Materials. Reagents and solvents were purchased from Aldrich and
used without further purification. The syntheses of the 4,4 -dicarboxylic
ng
0
Ψi ¼
cR,igR
ð1Þ
∑
R¼1
0
14
0
0
acid-2,2 -bipyridine (dcbpy) and 4,4 -bis(ethoxycarbonyl)-2,2 -bipyr-
15
idine (deebpy) ligands and of the [Ru(bpy)
deebpy) Cl ] 2H O complexes were carried out according to proce-
dures described in the literature.
Computational Details. General. All calculations were per-
formed with the ADF package. Geometrical optimizations were
2
symmetry using the PW91 exchange correlation
functional, which was selected among other functionals (LDA, BP86,
BLYP) to best reproduce the [Ru(bpy) ] crystalline structure. The
standard DZP basis set was employed for elements of the first series
C, N, O, H), while for ruthenium a TZP basis set was chosen.
2 2 2
Cl ] 2H O and [Ru-
3
with
(
2
2
2
3
16
ng
2
c
¼ 1 and ÆgRjgβæ ¼ 0 " R 6¼ β
ð2Þ
∑
R,i
17
R¼ 1
18
carried out under C
It follows that the probability of finding an electron on molecular orbital
is c ; in this sense, each MO can be characterized
R R
according to the fragment that prevails, corresponding to the fragment
1
9
20
2
Ψ
i
over fragment g
2
þ
3
21
2
R
with highest c .
21
(
On the other hand, a single excitation (γ) involving two molecular
orbitals can be defined. One of these MOs, Ψ , arises from the ground
state (GS) and the other, Ψ , from the excited state (ES). Accordingly,
TDDFT/ALDA calculations were also performed for the ligands and
H
22
metal complexes, employing the PBE exchange correlation and a
L
21
ZORA TZ2P basis set for ruthenium. A solvent effect study (acetonitrile)
was included only for complexes through the COSMO model where the
shape of the cavity is defined with a Klamt surface. Only the first 70
excitations were considered. The PBE exchange functional was selected
the process Ψ
H
f Ψ
L
will correspond to a displacement of electronic
} of Ψ to the fragment {g }of Ψ in
density from the fragment {g
R
H
β
L
accordance with the process γRβ: g f g or within the same fragment
R
β
g
R
f g
R
with R,β = {F,V,P,M,C,B,Bp}.
23
24
after checking others (LDA, BLYP, PW91, OLYP, LB94 ) for being
the one that best matches experimental spectra for the complexes under
study. Triplet energies were calculated using the ΔSCF method. In order
to interpret the emission bands, the difference in total energy was
determined in terms of two independent calculations, one for the energy
of the molecule at the optimized geometry in the first triplet state
configuration and the other for the molecule in the ground state,
although at the mentioned optimized geometry obtained for the triplet
state. All triplet state geometries were obtained through an unrestricted
calculation employing the PBE exchange functional. Optimized geometries
λ
Additionally, an electronic transition, Γ , can be visualized as the sum
of a series of single electronic excitations occurring at wavelength λ with
energy hν and probability of occurring measured by the oscillator
λ
strength f . Hence
nk
Γλ ¼ fλð
dk,λγRβ
Þ
ð3Þ
∑
k¼1
k
where dk,λ > 0 and ∑ dkλ = 1, which ensures that it is a one-photon
absorption process.
5
912
dx.doi.org/10.1021/ic1020862 |Inorg. Chem. 2011, 50, 5910–5924

Inorganic Chemistry
ARTICLE
Also, for the electronic transition occurring on the wavelength range
Hc). Anal. Calcd for C46
7.83. Found: C, 51.45; H, 3.89; N, 7.85.
[Ru(bpy) (LMe )](PF , 2. The synthetic procedure is identical to
that described for 1. Yield: 74%. IR, KBr (cm ): νCꢀH Aromatic = 3085,
H
36
N
6
RuF12
P
2
0.5H
2
O: C, 51.50; H, 3.48; N,
3
[
λ , λ ] it is possible to define a distribution function ζ(R,β) showing
i f
how a specific excitation, R f β, contributes along the overall interval of
2
2
6 2
)
ꢀ
1
wavelengths
1
!
ν
= 2920, ν
= 1364, ν
= 842, 558. H NMR
PꢀF
CꢀH Aliphatic
NꢀC
0 00
(CD ) CO) δ (ppm): 8.86 (s, 2H, H5), 8.77 (d, 4H, H2 , H2 ),
3 2
(
ζðR, βÞ ¼
fλðdk,λγRβÞ= Γλ for λ ∈ ½λi, λf ꢁ
ð4Þ
∑
∑
0
00
00
0
8
.17ꢀ8.12 (m, 6H, H3 , H3 , H5 ), 8.00 (d, 2H, H5 ), 7.74 (d, 2H,
λ
λ
0 00
H2), 7.63 (d, 2H, Hx), 7.58ꢀ7.48 (m, 10H, Ha, H3, H4 , H4 ), 7.07
Furthermore, the energy transferred from all transitions to the fragment
is calculated through
(
d, 2H, Hy), 6.74 (d, 4H, Hb), 2.99 (s, 12H, N(CH ) ). Anal. Calcd for
3 2
g
β
C
50 46 8 2 2
H N RuF12P H O: C, 51.42; H, 4.14; N, 9.59. Found: C, 51.71;
3
H, 4.21; N, 9.63.
6 2
Ru(deebpy) (LH)](PF ) , 3. A 100 mg (0.27 mmol) amount of LH
EðgβÞ ¼ EA ζðR, βÞ
ð5Þ
∑
[
2
β
was dissolved in 40 mL of hot benzene. In a second round flask, 214 mg
(0.27 mmol) of Ru(deebpy) Cl and 140 mg (0.55 mmol) of AgPF
6
where E
A
is the absorbed energy obtained from
EA ¼ fmaxΔν with Δν ¼ hc=ðλf ꢀ λiÞ
2
2
were dissolved in DMF. The latter mixture was added dropwise to the
LH ligand solution. Then, the mixture was heated to reflux for 6 h, and
then the system was allowed to cool to room temperature and the
solvent removed under vacuum. The formed solid was dissolved in
acetone and filtered on Celite. The filtrate was concentrated and
collected over cold hexane. The collected solid was filtered and washed
ð6Þ
Until now, the analysis has been mainly focused on the absorbed energy.
Nevertheless, by means of intersystem crossing (singlet to triplet) the
injection process may occur from different electronic states with respect
to those involved in the FranckꢀCondon absorption. Equation 7 permits
ꢀ
1
one to calculate I
fragment g into the T
T β
(g ), the amount of energy that can be delivered from
with ethyl ether. IR, KBr (cm ): νCꢀH Aromatic = 3000ꢀ3083, νCdO
=
1
0
00
β
1
state
1728, ν PꢀF = 839. H NMR (CD
3
CN) δ (ppm) 9.07 (d, 2H, H2 , H2 ),
0
00
8
.70 (s, 1H, H5), 8.02 (d, 1H, H2), 7.91ꢀ7.83 (m, 2H, H3 , H3 ), 7.78
ITðgβÞ ¼ EðgβÞj ðgβÞ
ð7Þ
0 00
T
(
d, 1H, Hx), 7.68 (d, 2H, H5 , H5 ), 7.54 (d, 2H, Ha), 7.46 (t, 3H, Hb,
Hc), 7.41 (d, 1H, H3), 7.31 (d, 1H, Hy), 4.44 (q, 4H, OCH CH ), 1.38
2
3
In this equation E(g ) is calculated from eq 5 and j (g ) is a distribution
function with a maximum value of one, which contains information about
β
T
β
(
3
PF
t, 6H, OCH
2
CH
3
). Anal. Calcd for C58
H
52
N
6
O
8
RuF12
P
2
: C, 51.52; H,
þ
.88; N, 6.22. Found: C, 51.39; H, 3.90; N, 6.20. m/z: 1208 (M
ꢀ
the participation of the different fragments in the state under study. Data
ꢀ
þ
ꢀ
2
5
); 531 (M ꢀ 2PF
6
).
for this function is obtained through a fragment population analysis.
6
0
0
2 2 6 2
[Ru(deebpy) (LNMe )](PF ) , 4. The synthetic procedure was iden-
Synthesis. 4,4 -Bis(2-hydroxy-2-phenyl)ethyl-2,2 -bipyridine and
ꢀ
1
0
0
tical to that described for 3. Yield: 70%. IR, KBr (cm ): ν
3
5
=
4
,4 -Bis[2-hydroxy-2-p-(dimethylaminophenyl)ethyl]-2,2 -bipyridine.
2
CꢀH Aromatic
6
121, νCꢀH Aliphatic = 2863, νCdO = 1727, νCꢀN = 1384, ν PꢀF = 840,
These syntheses were carried out as described in the literature using
1
0
00
58. H NMR ((CD ) CO) δ (ppm): 9.27 (s, 2H, H5 , H5 ), 8.92
benzaldehyde and p-toluoldehyde, respectively.
3 2
0
0
(s, 1H, H5), 8.40 (d, 1H, H2), 8.33 (d, 2H, Hb), 8.00 (d, 1H, H3), 7.90
4
,4 -Distyryl-2,2 -bipyridine, LH. A 602 g (1.52 mmol) amount of the
0 00
d, 2H, Ha), 7.69 (d, 1H, Hx), 7.50 (d, 1H, H2 ), 7.44 (d, 1H, H2 ), 7.07
(
(
diol and 60 mL of acetic acid are combined. The resulting mixture is
heated to reflux for 18 h. Later, the solution is allowed to cool, preventing
the acetic acid from freezing. The formed solid is separated by filtration
0
00
d, 1H, Hy), 6.72 (d, 1H, H3 ), 6.63 (d, 1H, H3 ), 4.41 (m, 4H,
CH ), 2.98 (s, 6H, N(CH ), 1.33 (t, 6H, OCH CH ). Anal.
OCH
2
3
3
)
2
2
3
ꢀ1
Calcd for C H N O RuF P 0.7H O: C, 51.33; H, 4.4; N, 7.72.
and washed with ethyl ether. Yield: 65%. IR, KBr (cm ): νCꢀH Aromatic
62 62
8
8
12
2
3
2
1
Found: C, 51.32; H, 4.48; N, 7.75.
Ru(LH) Cl ] and [Ru(LNMe
=
3027. H NMR (CDCl
3
) δ (ppm): 8.69 (d, 1H, H2), 8.55 (s, 1H, H5),
[
2
2
2
)
2
Cl
2
]. This synthetic procedure is
7.58 (d, 2H, Ha), 7.47 (d, 1H, Hx), 7.40 (t, 3H, Hb, Hc), 7.34 (d, 1H,
1
6
carried out similarly to that described by Sullivan, mixing 250 mg
0.69 mmol) of ligand LH, 71 mg (0.27 mmol) of RuCl 3H O, 38 mg
H3), 7.15 (d, 1H, Hy). Anal. Calcd for C26
H
20
N
2
: C, 86.64; H, 5.59; N,
(
3
2
7
.77. Found: C, 86.39; H, 5.63; N, 7.74.
3
0
0
4
,4 -Bis[p-(dimethylamino)-R-styryl]-2,2 -bipyridine, LNMe . The
(0.36 mmol) of hydroquinone, and 470 mg (11.1 mmol) of LiCl in
100 mL of absolute ethanol and then heating to reflux for 8 h. Afterward,
the mixture is allowed to cool to room temperature. The remaining solid
is filtered under vacuum and washed with small portions of ethyl ether.
2
2
6
synthetic procedure is carried out just as described by A. Juris. IR,
ꢀ1
1
KBr (cm ): νCꢀH Aromatic = 3021, νCꢀN = 1360. H NMR (CDCl
3
) δ
(ppm): 8.60 (d, J = 5.1 Hz, 1H, H2), 8.46 (s, 1H, H5), 7.45 (d, J = 8.8 Hz,
1
2
2
1
2
H, Ha), 7.39 (d, J = 16.2 Hz, 1H, Hx), 7.33 (dd, J = 5.1 Hz, J = 1.6 Hz,
For complex [Ru(LNMe
950 mg (2.13 mmol) of ligand LNMe
RuCl 3H O, 140 mg (1.29 mmol) of hydroquinone, and 450 mg
(10.6 mmol) of LiCl.
[Ru(deebpy)(LH) ](PF
is added to 20 mL of methanol. In a second round flask, 200 mg (0.22
mmol) of [Ru(LH) Cl ] and 113 mg (0.44 mmol) of AgPF in 25 mL of
2
)
2
Cl
2
] the same procedure is carried out, using
H, H3), 6.91 (d, J = 16.2 Hz, 1H, Hy), 6.71 (d, J = 8.34 Hz, 2H, Hb),
2
, 280 mg (1.06 mmol) of
.99 (s, 6H, Nꢀ(CH
3
)
2
). Anal. Calcd for C30
H
30
N
4
0.5H
2
O: C, 79.07;
3
2
3
3
H, 6.86; N, 12.29. Found: C, 79.07; H, 6.89; N, 12.38.
Ru(bpy) (LH)](PF ) , 1. The synthetic procedure is carried out as
described by B. P. Sullivan. A 69.3 mg (0.192 mmol) amount of LH
and 97.2 mg (0.384 mmol) of AgPF are dissolved in 25 mL of ethanol.
To the resulting mixture, 100 mg (0.192 mmol) of Ru(bpy) Cl 2H O
[
2
6 2
) , 5. A 67 mg (0.22 mmol) amount of deebpy
2
6 2
27
6
2
2
6
methanol are added dropwise to the deebpy solution. The mixture is
heated to reflux for 3 h. Later, the mixture is allowed to reach room
temperature and all the solvent is evaporated. The resulting solid is
dissolved in acetone and filtered on Celite. The filtrated portion is
concentrated and added dropwise to ethyl ether. The resulting solid is
removed by filtration under vacuum and washed with ethyl ether. IR,
2
2
3
2
is added. Then, the mixture is heated to reflux for 6 h, shielding the
system from light by aluminum foil. Afterward, the mixture is filtered
4 6
over Celite and a stoichiometic quantity of NH PF is added to the
remaining fluid. The formed solid is filtered under vacuum and purified
by column chromatography over aluminum oxide using ethanol fol-
lowed by acetone. The acetone portion is concentrated and collected on
ꢀ
1
1
KBr (cm ): νCꢀH Aromatic = 3028ꢀ3057, νCdO = 1727, ν PꢀF = 840. H
NMR (CD CN) δ (ppm): 9.03 (s, 1H, H5 ), 8.74 (d, 2H, H2, H2 ),
8.04 (d, 1H, H2 ), 7.83 (d, 1H, H3 ), 7.75 (d, 2H, Hx , Hx ), 7.73 (s, 2H,
ꢀ
1
0
00
ethyl ether. Yield: 68%. IR, KBr (cm ): ν
= 3082, ν
=
3
CꢀH Aromatic
PꢀF
1
0
0
0
00
8
(
(
40, 557. H NMR ((CD ) CO) δ (ppm): 9.05 (s, 2H, H5), 8.79
3
2
0
00
0
00
00
00
0
00
00
0
d, 4H, H2 , H2 ), 8.18 (t, 4H, H3 , H3 ), 8.14 (d, 2H, H5 ), 8.03
H5, H5 ), 7.67 (d, 4H, Ha , Ha ), 7.53ꢀ7.38 (m, 8H, H3, H3 , Hb ,
0
00 0 00 0 00
d, 2H, H5 ), 7.90 (d, 2H, H2), 7.80 (d, 2H, Hx), 7.70ꢀ7.68 (m, 6H, H3,
Hb , Hc , Hc ), 7.30 (d, 2H, Hy , Hy ), 4.42 (q, 2H, OCH
2 3
CH ), 1.39
0
00
Ha), 7.56 (dd, 4H, H4 , H4 ), 7.43 (d, 2H, Hy), 7.42ꢀ7.34 (m, 6H, Hb,
(t, 3H, OCH CH ). Anal. Calcd for C H N O RuF P : C, 57.83;
2
3
68 56
6
4
12 2
5
913
dx.doi.org/10.1021/ic1020862 |Inorg. Chem. 2011, 50, 5910–5924

Inorganic Chemistry
ARTICLE
1
Figure 2. H NMR spectra for ligands LH and LNMe
2
in (CDCl
3
) and complexes 1 and 2 in ((CD
3
)
2
CO).
þ
H, 4.00; N, 5.95. Found: C, 57.63; H, 4.03; N, 5.93. m/z: 1268.8 (M ꢀ
chromophoric ligand. Specifically, the position of the two signals
assigned to the Hx and Hy vinyl protons (see proton labeling
in Scheme 1) is sensitive to the presence of the amino groups.
A chemical shift to higher field is observed for ligand LNMe2,
with a difference of 0.08 ppm for Hx and 0.24 ppm for Hy with
respect to analogous signals from these protons in the LH ligand
spectrum. Similarly, in complex 2 the signals of the Hx and Hy
protons are shifted to higher field compared to complex 1. The
shift of the signals is more significant for the complexes than for
the ligands. The higher shielding of the protons in the vinyl (V)
ꢀ
þ
ꢀ
6
PF
6
); 562 (M ꢀ 2PF
)
[
Ru(deebpy)(LNMe
2
)
2
](PF
6
)
2
, 6. An 85 mg (283 mmol) amount of
deebpy is dissolved in 10 mL of DMF, and a suspension of 300 mg
282 mmol) of [Ru(LNMe Cl ] in 40 mL of DMF is added. To the
resulting mixture, 142 mg (562 mmol) of AgPF is added. The mixture is
heated to reflux for 6 h and protected from light. Then, the hot mixture is
filtered through Celite and the filtrate solution evaporated. The remain-
ing solid is dissolved in acetone and precipitated in ethyl ether. Yield:
(
2
)
2
2
6
ꢀ
1
7
4%. IR, KBr (cm ): νCꢀH Aromatic = 3026, νCꢀH Aliphatic = 2987, 2923,
1
2
853, νCdO = 1723, ν
= 1365, ν
= 840, 558. H NMR (CD CN)
PꢀF 3
NꢀC
fragment for LNMe and its complexes, compared to LH,
2
δ (ppm): 8.99 (s, 2H), 8.62 (s, 3H,), 8.06 (t, 2H), 7.86 (d, 2H), 7.73
m, 5H), 7.57 (m, 7H), 7.31 (m, 11H), 6.92 (m, 10H), 4.42 (m, 4H,
OCH CH ), 3.01 (s, 24H, N(CH ), 1.38 (t, 6H, OCH CH ). Anal.
Calcd for C H N O RuF P H O: C, 56.96; H, 4.91; N, 8.74.
correlates well with a higher percentage of electronic density in
this fragment in the HOMO orbital, as discussed below (see, e.g.,
UVꢀvis spectra section).
(
2
3
3
)
2
2
3
7
6
76 10
4
12
2
3
2
Figure 3 summarizes the results of molecular orbital calcula-
tions displayed from Tables S1ꢀS6 (Supporting Information).
The information is shown by means of molecular orbital energy
diagrams for each of the ruthenium complexes under study.
A rather different behavior is observed for complexes containing
the LH ligand (1, 3, and 5) when compared to analogous
Found: C, 57.13; H, 5.10; N, 8.71.
’
RESULTS AND DISCUSSION
The IR spectra, taken in KBr, have characteristic bands for the
carboxylic groups between 1719 and 1730 cm as well as bands
belonging to the PF6 ion at 558 and 840 cm . For complexes 2,
, and 6, a band at 1364 cm is assigned to the stretching of the
ꢀ
ꢀ1
1
ꢀ
complexes with the LNMe ligand (2, 4, and 6). For example,
2
ꢀ
1
4
in complex 1 the five occupied molecular orbitals of higher energy
contain pure d metal orbitals and mixtures of d orbitals with
π orbitals from the chromophoric ligand. The first six unoccupied
molecular orbitals are dominated by those arising from the
LUMOs of the bipyridine ligands and the chromophoric ligand.
Electronic density analysis shows that the LUMO is localized
on the P fragment (Figure 1) of the chromophoric ligand; the
LUMOþ1 is found 0.11 eV higher in energy and is localized
in the bipyridine ligand. As expected, the presence of an
CꢀN bond of the donor NMe group. The elemental analyses
2
for these compounds are in agreement with the proposed
structures. In compounds such as 3 and 5, where some deviation
was observed, mass spectra permitted confirming the presence of
the proposed complex.
1
Electronic Density Distribution. The H NMR spectra for
ligands LH and LNMe and complexes 1 and 2, Figure 2, show
the electron-donor character of the NMe substituent in the
2
2
5
914
dx.doi.org/10.1021/ic1020862 |Inorg. Chem. 2011, 50, 5910–5924

Inorganic Chemistry
ARTICLE
Figure 3. Molecular orbital diagram for ruthenium complexes (Ru = ruthenium; B = bipyridine; Bp = dimethyl dicarboxylate of bipyridine; LH = 4,
0
0
0
0
4
-distyryl-2,2 -dipyridine; LNMe = 4,4 -bis[p-(dimethylamino)-R-styryl]-2,2 -bipyridine).
2
electron-withdrawing group, such as ꢀCOOMe, in the bipyridine
ligand causes a lowering in the energy of the levels, mainly in the
MOwherethisligandparticipates. As a consequence, in complexes
with no amino substitution. Table 1 summarizes the preceding
analysis. As mentioned, the effect of the donor NMe substitu-
2
ents on the choromophoric ligand and that of acceptor carbox-
ylate substituents on the bpy ligands is reflected in the nature of
the HOMO on the nature of the LUMO and on the energy
difference between them, ΔE.
3
and 5 the LUMO is centered on the bipyridine orbitals, Bp, with
a lower energy than that in the chromophoric ligand orbitals, LH.
On the other hand, when the electron-donating dimethylamino
substituent is introduced in the para position of the phenyl ring of
The electrochemical experimental results for the complexes
show consistency with the theoretical MO calculations. Table 2
shows the oxidationꢀreduction potentials for the series of com-
plexes, measured by cyclic voltammetry. It is noteworthy that a
correlation coefficient of 0.95 is found when comparing the first
electrochemical oxidation potential with the calculated energy of
the HOMO orbital for the series of complexes under study. This
the chromophoric ligand, as in ligand LNMe , the opposite effect
2
is observed: all levels show an increase in energy. The effect is
particularly notable for complex 2, while it seems to be attenuated
in complexes 4 and 6, where an anchoring carboxylate acceptor
group is present on the bipyridine ligands.
Regarding the HOMO for complex 1, it shows two nearly
degenerated orbitals: one centered on the metal, resulting from a
28
relationship emerges from KohnꢀSham theory, which states in
29
combination of d orbitals (d þ d
þ d ), and the other a
its exact formulation that the frontier eigenvalue is the ioniza-
30
xy
x2ꢀy2
z2
mixture of the d þ d orbitals with a π orbital from the
tion potential. However, since solvent effects are included in
the present calculations, a correlation between the HOMO
eigenvalue and the first oxidation potential is expected. This
relationship opens the possibility to employ theoretical informa-
tion to gain insight on the trend of electrochemical experimental
results.
xz
yz
chromophoric ligand. A similar composition is also found for all
complexes containing the LH ligand and is independent of the
substituent in the bipyridine ligand. The energy of the corre-
sponding HOMOs varies between ꢀ6.0 and ꢀ6.5 eV. However,
when the dimethylamino substituent is introduced, the energy of
the HOMO in complexes 2, 4, and 6 is raised above ꢀ5.8 eV
Specifically, looking at the experimental data in Table 2, it can
be observed that an irreversible oxidation occurs at 449, 488, and
459 mV for complexes 2, 4, and 6, respectively. These potentials
are absent in compounds 1, 3, and 5 and are therefore attributed
(Tables S2, S4, and S6 in the Supporting Information). This
value is the result of the competition between the electron-
attracting effect of the substituted bipyridines and the effect of
the electron-donor substituent on the chromophoric ligand. The
electron density distribution of the HOMO is also altered, being
mainly located on the F fragment of the chromophoric ligand,
while the HOMOꢀLUMO band gap of these complexes is
reduced about 40% with respect to the corresponding complexes
13a
to oxidation of the amino group. From the MO results in
Figure 3 and Table 1, this assumption can be corroborated as
complexes 2, 4, and 6 possess a HOMO orbital, related to the first
oxidation, centered on the dimethylamino group. The same
result is obtained by a fragment analysis performed in the same
5
915
dx.doi.org/10.1021/ic1020862 |Inorg. Chem. 2011, 50, 5910–5924

Inorganic Chemistry
ARTICLE
a
Table 1. Substituent Effect on the Energy of Frontier Molecular Orbitals for Complexes 1ꢀ6
complex
1
2
3
4
5
6
L (chromophoric ligand)
R (for R-bpy)
LH
H
LNMe
2
LH
LNMe
2
LH
LNMe
2
H
COOEt
Ru-L
bpy
COOEt
L
COOEt
Ru
COOEt
L
HOMO centered on
LUMO centered on
Δ(HOMOꢀLUMO) (eV)
Ru
L
L
bpy
bpy
bpy
bpy
1.78 1.22
1.62
1.05
1.51
1.03
b
overall substituent effect
HOMO and LUMO HOMO and LUMO HOMO destabilized HOMO and LUMO HOMO destabilized
destabilized stabilized LUMO stabilized stabilized LUMO stabilized
on the Δ(HOMOꢀLUMO) gap
a
b
Substituents: Donor NMe group on the chromophoric ligand and COOEt the acceptor group on bpy. Referred to the fully unsubstituted complex 1.
2
a
Table 2. Electrochemical Data of the Complexes
Finally, complexes 5 and 6 can be visualized as originating
from the “replacement” in complexes 3 and 4 of a ligand with an
2
þ/3þ
2þ/3þ
E
pa
E
1/2 (Ru
)
ΔE
p
(Ru
)
acceptor substituent, (bpy-(COOH) ), by a chromophoric li-
2
b
c
d
complexes
(mV)
(mV)
(mV)
gand. Both facts (the decrease of the number of acceptor ligands
and the increase in the number of chromophoric ligands) should
tend to make the oxidation easier, and this is clearly observed
both in the experimental electrochemical data as well as in the
MO calculations.
1
2
3
4
5
6
931
59
449
488
459
1013
1115
1151
968
145
89
282
55
UVꢀVis Spectra Analysis. 1. Ligands. Absorption spectra of
ligands LH and LNMe show a broad and intense band centered
e
2
1232
at 317 and 390 nm, respectively, as can be seen in Figure 4, where
results from TDDFT calculations on the gas phase are also
displayed. The absorption bands of ligands LH and LNMe2
appear displaced to lower energies when compared to bipyridine
(λmax = 280 nm). This red shift has been attributed to the
a
ꢀ4
Cyclic voltammograms were taken in a 1 ꢂ 10 M solution of the complex
at a sweep rete of 100 mV/s in dry CH CN containing 0.1 M tetrabuty-
3
lammonium hexaflourophosphate as supporting electrolyte (working elec-
trode, disc Pt; reference electrode, Ag/Ag , and counter electrode, Pt).
þ
b
c
d
e
Oxidation of substituents. E1/2 =1/2(Epa þ Epc). ΔE
p
=|Ep,c ꢀ Ep,a|. Epc
.
31
presence of the aromatic π-conjugated system. The effect is
enhanced in the LNMe ligand, due to the presence of the
way as described before for the free ligand. The second oxidation
potential in these complexes is centered on the metal and appears
at values similar to those from the first oxidation potential of
complexes 1, 3, and 5.
2
electron-donating aminoalkyl substituent, Figure 4b. It has also
been suggested that these bands correspond to intraligand charge
transfer, ILCT, transitions. Addition of a small amount of acid to a
solution of the ligand diminishes the intensity of the 380 nm band,
probably due to the protonation of the amino group, which,
therefore, can no longer participate in the ILCT transition. This
experimental test supports the ILCT nature of the bands.
2þ/3þ
From the ΔE (Ru
) values of Table 2, it can be inferred
p
that complexes 1, 3, and 5 exhibit a reversible oxidation process;
moreover, experimentally these complexes exhibit a i /i ratio
pc pa
close to unity. Conversely, for complexes 2 and 4 enhanced ΔEp
values are observed, reflecting a more irreversible behavior, while
complex 6 is definitely irreversible. The tendency of complexes 2
and 4 to be more irreversible, with respect to 1 and 3, can
be related to their lower thermodynamic stability. Regarding
the metal oxidation potential, an expected trend is observed
when the electron-withdrawing carboxylate groups are present,
Molecular orbital analysis in terms of the fragment compo-
nents displayed in Table 3 shows that for the ligand LH the
HOMO possesses σ character resulting from the mixing of p
þ s
x
atomic orbitals in the P fragment and that the HOMOꢀ1 and
HOMOꢀ2 molecular orbitals possess π character and a more
delocalized electronic distribution, with a high contribution of
the V and F fragments. On the other hand, the LUMO and
LUMOþ1 orbitals exhibit π-antibonding character delocalized
over the whole molecule, with the main electronic occupancy
on the P fragment for the LUMO and V and F fragments for
LUMOþ1. Finally, LUMOþ2 is again dominated by the P
2
þ/3þ
reflected in a more difficult Ru
oxidation for complex 3
(which possesses electron-withdrawing groups, deebpy, in both
bpy ligands) compared to 1 and for complex 4 compared to 2.
Specifically, for complexes 1 and 3, values of 931 and 1115 mV,
respectively, are reported. Similarly, complexes 2 and 4 follow the
2
þ/3þ
same trend, with E1/2 (Ru
) values of 1013 and 1151 mV,
fragment. Regarding the LNMe ligand, where the dimethyla-
2
respectively. This reflects the expected fact that the bipyridinic
ligands with acceptor groups remove electronic density from the
ruthenium metallic center, making oxidation of the metal a more
energy-demanding process.
mino substituent has been introduced, the molecular orbital
fragment analysis shows, Table 4, that the first three LUMOs are
of composition similar to those from the LH ligand. However, a
difference is observed in the nature of the HOMO and HOMOꢀ1,
The effect of the donor NMe group in the chromophoric
where LNMe has a composition dominated by the F fragment
2
2
ligand on the metal oxidation is less clear. The presence of the
donor substituent should make the metal oxidation easier. The
experimental trend in Table 2 goes in the opposite direction.
Nevertheless, the behavior can be explained by the fact that
the ruthenium oxidation occurs on an already positively charged
species. Otherwise, it can also be concluded that when both
donor and acceptor substituents are present the effect of the
latter seems to predominate.
and a nearly degenerate character for these orbitals.
Additionally, TDDFT calculations were performed, and the
results displayed in Table 5. These types of calculations help
to obtain insight into the electronic transitions responsible for
the absorption spectra. Two main transitions (1A and 1B in
Figure 4) having high oscillator strength (0.56 and 0.94,
respectively) are responsible for the large absorbance found
for ligand LH at 317 nm. Both transitions contain two types of
5
916
dx.doi.org/10.1021/ic1020862 |Inorg. Chem. 2011, 50, 5910–5924

Inorganic Chemistry
ARTICLE
Figure 4. Comparison of UVꢀvis experimental spectra in CH
3 2
CN with TDDFT results for ligands (a) LH and (b) LNMe .
Table 3. Decomposition of Molecular Orbitals in Terms of
Orthogonal Fragments for LH
on the π orbital of the FV fragment. Figure 5a shows in a
pictorial way an IL excitation, which would be predominant in
the mentioned LH bands.
fragment
Regarding the LNMe ligand, the HOMO and HOMOꢀ1
2
degenerate molecular orbitals centered in the F fragment parti-
cipate in transitions 2B and 2C, which result in the red-shifted
absorption appearing at 390 nm for the intraligand ILCT
transitions from fragment F to fragment P, Figure 5c.
MO
occupation
energy
F
V
P
assignment
LUMOþ2
LUMOþ1
LUMO
0.0
0.0
0.0
2.0
2.0
2.0
ꢀ2.70
ꢀ3.28
ꢀ3.57
ꢀ6.08
ꢀ6.22
ꢀ6.28
9
3
88
28
50
100
27
23
π*(P)
35
22
0
37
28
0
π*(FþV)
π*(P)
2
. Complexes. The absorption spectra for the ruthenium
complexes studied in this work show two characteristic bands:
The first one appears in the UV region and shows high ε values.
This band can be associated with the absorbance of the chro-
mophoric ligand due to its similarity to the band appearing in the
absorption spectra of the uncoordinated ligands, although red
shifted due to the effect of coordination to the metallic center.
The second broad and less intense band is located in the visible
HOMO
σ(P)
HOMOꢀ1
HOMOꢀ2
37
41
36
36
π(FþV)
π(FþV)
Table 4. Decomposition of Molecular Orbitals in Terms of
Orthogonal Fragments for LNMe2
32
region and, according to literature, is mainly assigned as a
MLCT band, considering also the fact that it appears at lower
energy. Regarding solar cell devices, the second band is more
relevant in the injection process and therefore is interesting to
analyze. Also, since the energy absorbed comes from the visible
region, it could be expected that modifications of the structure of
the complexes will result in a change of the energy and intensity
of the band, giving rise to an improvement of the solar cell device.
In order to gain more insight into the composition of this broad
band, a transition density analysis (TDA), coupled with a
TDDFT calculation, was performed over the series of complexes
under study, in a similar way to that done for the ligands.
However, for the case of the complexes, molecular orbitals were
decomposed in five basic fragments containing ruthenium,
bipyridine, and the three F, V, P fragments coming from the
chromophoric ligands. According to this and considering the
fragment
MO
occupation
energy
F
V
P
assignment
LUMOþ2
LUMOþ1
LUMO
0.0
0.0
0.0
2.0
2.0
2.0
ꢀ2.04
ꢀ2.50
ꢀ2.84
ꢀ4.94
ꢀ4.97
ꢀ5.43
8
5
87
31
54
12
11
100
π*(P)
π*(FþV)
π*(P)
π(F)
32
20
70
71
0
37
26
18
18
0
HOMO
HOMOꢀ1
HOMOꢀ2
π(F)
σ(P)
excitations: intraligand (IL), which takes place parallel to the
molecular axes and involves an electronic displacement from
the FV fragment to the P fragment, and ππ*, with an electronic
displacement perpendicular to the molecular axes and centered
5
917
dx.doi.org/10.1021/ic1020862 |Inorg. Chem. 2011, 50, 5910–5924

Inorganic Chemistry
ARTICLE
Table 5. Assignment of the Main Contributing Transitions to the Overall Absorption Bands in LH and LNMe2
ligand
LH
transition
wavelength (nm)
oscillator strength
fragments involved
assignment
1A
311
329
386
403
0.56
0.94
0.63
0.53
FV f P þ FV f FV
FV f P þ FV f FV
F f P
IL þ ππ*
IL þ ππ*
ILCT
1
B
LNMe
2
2B
2
C
F f P
ILCT
Figure 5. Example of electronic transitions in free ligands involving two molecular orbitals described in terms of orthogonal fragments: (a) intraligand
IL); (b) π f π*; (c) intraligand charge transfer (ILCT). Note that ILCT bands imply a remarkable charge displacement from one region of the ligand
(
to another, with a consequent change in dipolar moment.
predominance of each fragment in the overall composition,
molecular orbitals can be classified as centered on the metal,
centered on the chromophoric ligand, or centered on the
bipyridine or mixtures between them. On the other hand, the
TDA method shows that each transition is composed by a sum of
the single electronic transitions that link an occupied MO from
the ground state (GS) with a corresponding MO from the excited
state (ES). Employing the above classification of the MO it is
possible to distinguish four basic types of electronic transitions,
which are depicted in Figure 6. Electronic transitions a and b
belong to the category of metal-to-ligand charge transfer, MLCT,
and involve a charge transfer from the ruthenium atom to the
chromophoric ligand (MC) or to bipyridine (MB). The ligand to
ligand (LL) transition, shown in c, corresponds to a charge
transfer between the F-substituted fragment of the chromophoric
ligand and bipyridine. Finally, Figure 6d shows a charge transfer
Furthermore, the total height of each column corresponds to the
oscillator strength of the corresponding transition. Hence, in this
way it is easy to determine how the calculated transitions
influence the shape and maxima of the experimental band.
According to Figure 7, the spectra of complex 3 with λ
=
max
ꢀ
1
ꢀ1
473 nm and εmax = 24 000 M cm contain seven main
transitions located within the range from 410 to 580 nm.
Transition 3D occurs at 457 nm, has the highest oscillator
strength (0.29), and contains mainly an IL electron transfer
process. The metal to bipyridine electron transfer (MB) can be
found contributing to transitions 3C (449 nm), 3E (461 nm),
and 3F (484 nm), although for the latter two the MB process is
coupled with IL and LL basic transitions. When the number of
“chromophoric ligands” present in the complex is increased, as is
the case of complex 5 with respect to 3, the extinction molar
ꢀ1
ꢀ1
coefficient increases from 24 000 to 35 000 M cm . The
simulated spectrum of 5 shows, in the range from 450 to
550 nm, an increase in the number of transitions corresponding
to MLCT bands. Moreover, transition 5J, located at 494 nm and
process that occurs inside the chromophoric LNMe ligand,
2
32
which is classified as an intraligand process (ILCT). If the
intraligand process does not show clear charge transfer character,
it is named as IL, as would be the case of complexes with the LH
ligand.
showing the highest oscillator strength (0.19), is close to λmax
.
The reduced contribution of IL and LL basic transitions to the
overall absorption of this complex is noteworthy. On the other
hand, the presence of the dimethylamino substituent on the
Figure 7 shows the experimental spectra for complexes 3ꢀ6 as
well as the single transitions calculated by TDDFT in terms of
stacked columns representing the contribution of each of the
four basic electronic transitions to the overall absorption band.
chromophoric ligand (LNMe ligand) produces a significant
2
increase in the ε value, as is the case of complex 4, with a εmax of
5
918
dx.doi.org/10.1021/ic1020862 |Inorg. Chem. 2011, 50, 5910–5924

Inorganic Chemistry
ARTICLE
Figure 6. Example of basic electronic transitions in ruthenium complexes involving two molecular orbitals described in terms of orthogonal fragments:
a) MC = metal to chromophoric ligand charge transfer; (b) MB = metal to bipyridine ligand charge transfer; (c) LL = interligand transfer from
chromophore to bipyridine; (d) ILCT = intraligand charge transfer.
(
ꢀ
1
ꢀ1
ꢀ1
ꢀ1
4
6 800 M cm , compared to 24 000 M cm of complex 3.
of the metal to chromophore electronic transitions. On the other
hand, when two chromophoric ligands are present (complex 5)
a predominance of the metal to chromophore transition is
observed. Taking again as reference the model complex 1,
introduction of the donor amino substituent in the chromo-
phore, as in 2, causes a detrimental effect over MC transitions and
favors LL and MB transitions. Attaching the methyl ester
electron-acceptor substituent, as in 4, results mainly in the
enhancement of metal to bipyridine charge transfer transitions.
The intraligand charge transfer process (ILCT) is noticeably
favored when two electron-donor-substituted chromophoric
The TDDFT calculations show the appearance of two very
intense transitions, 4F and 4G, located at 479 and 489 nm with
oscillator strengths of 0.46 and 0.42, respectively. Both 4F and
G are composed by a mixture of MB, LL, and ILCT electronic
transitions, with predominance of the latter. Transitions 4J and
K, mainly composed of ILCT and LL transitions, are respon-
sible for the increase of the absorbance on the tail of the band and
the enhancement of its width up to 570 nm. On the other hand,
complex 6, with two chromophoric LNMe ligands, exhibits an
4
4
2
increase of the extinction molar coefficient up to 67 000
ꢀ1
ꢀ1
12
M
cm and the appearance of a maximum at 426 nm.
ligands are present, as in 6, in accordance with previous reports.
Theoretical analysis reveals an increase in the number of transi-
tions to be considered (from 8 to 11) and the presence of two
transitions located at 407 (6A) and 524 nm (6I) with oscillator
strengths of 0.26 and 0.79, respectively. The 6A transition is
mainly a MLCT electron transition from ruthenium to the
chromophoric ligand, while 6H and the neighboring bands 6J,
By combining the results from Figures 7 and 8, it is possible to
build a simple theoretical model predicting the behavior of such
complexes when employed as dyes in solar cell devices. The
white column in Figure 9 represents the total energy absorbed
(E ) by the complexes in the range from 0.113 to 0.076 eV
A
(400ꢀ600 nm). This value is calculated through the product
6
I, and 6 L evidence a predominance of ILCT character.
f
Δν, eq 6, which is proportional to the area under the curve of
max
The overall contribution of the four basic electronic transitions
the experimental absorption spectra (fmax is the highest oscilla-
tor strength value obtained from TDDFT). From this figure it
can be seen that, as expected, the amount of energy absorbed is
correlated with the number of chromophoric ligands present in
the complex. On the other hand, the increase in absorbed energy
is evident when chromophoric ligands with dimethylamino
substituents are present in the complex, as in 4 and 6, when
compared to 3 and 5, respectively. Additionally, the black and
striped columns in Figure 9 show the amount of absorbed
in the absorption region comprised between 400 and 600 nm for
the series of complexes under study is displayed in Figure 8.
As can be seen, each complex possesses a characteristic pattern
determined by the acceptorꢀdonor character of the substituents
on both ligands. For the unsubstituted model complex 1 the
absorption band is composed mainly by intraligand (IL) and
metal to chromophoric ligand (MC) processes. When an elec-
tron-attracting substituent, such as ꢀCOOMe, is attached to the
bipyridine ligand (complex 3), an enhancement of the electronic
transfer between metal to bipyridine (MB) occurs at the expense
energy delivered to the bypiridinic, E (B), and chromophoric,
D
ED(C), ligands by means of (MB þ LL) and MC absorption
5
919
dx.doi.org/10.1021/ic1020862 |Inorg. Chem. 2011, 50, 5910–5924

Inorganic Chemistry
ARTICLE
Figure 7. Comparison of UVꢀvis experimental spectra in CH
3
CN with TDDFT results (Γ
λ
function of eq 3) for the ligands in complexes 3ꢀ6.
Figure 8. Distribution function, ζ(R,β) in eq 4, of the four basic types of electronic transitions along the visible band (400ꢀ600 nm) for complexes 1ꢀ6.
5
920
dx.doi.org/10.1021/ic1020862 |Inorg. Chem. 2011, 50, 5910–5924

Inorganic Chemistry
ARTICLE
Figure 9. Available energy for electronic injection; E = absorbed energy, E (B) = energy delivered to bipyridine, E (C) = energy delivered to the
A
D
D
chromophoric ligand.
processes, respectively. For an efficient direct electronic injec-
tion, most of the absorbed energy should be transferred to the
ligand anchored on the semiconductor surface. However, in-
spection of Figure 9 shows that the greater amount of absorbed
energy (white column) in 5 and 6, with respect to 3, does not
necessarily mean an increase in the amount of direct electronic
injection capacity (black column). From the E /E ratio
Table 6. Spectroscopic Data for Ligands and Complexes
a
ꢀ1
ꢀ1
b
compound
LH
λ
max(Abs) (nm)
ε (M cm
)
λ
max(Em) (nm)
317
390
465
ILCT
376
520
642
LNMe
2
ILCT
1
MLCT
ILCT
22 300
46 200
46 900
42 000
24 000
43 500
46 800
40 900
35 000
92 000
63 400
67 000
D
A
3
26
468
10
473
49
(
Tables S3ꢀS6 in the Supporting Information) it can be
2
3
4
5
6
MLCT
ILCT-ML
MLCT
ILCT
673
629
estimated that for complexes 5 and 6 a maximum of 14% of
the absorbed energy is available for injection (i.e., only 14% of
the electronic density after excitation is concentrated on the
anchoring ligand), compared with the 51% available for 4 and
4
3
477
370
486
329
480
MLCT
ILCT
3
4% for 3.
Emission and Triplet State. The chromophoric LH and
LNMe ligands have an emission band in acetonitrile at room
MLCT
ILCT
664
2
temperature. Table 6 summarizes the relevant emission data for
these ligands and their corresponding complexes. Complexes 1,
MLCT
ILCT
2
, 3, and 5 have emission processes in acetonitrile at room
426
temperature in the absence of oxygen. It should be mentioned
that the electrochemical data of Table 2, specifically the differ-
ence Δ between the first reduction and the first oxidation
potentials, show a good correlation with the emission results.
For example, the Δ value diminishes from 2.35 to 2.26 V on going
from 3 to 5 while EEM decreases from 1.97 to 1.87 eV.
a
b
3 3
Solvent: CH CN, Determined at room temperature in CH CN.
predominance of nonradiative deactivation pathways or to a sort
of quenching process by the ester group.
At this point it is relevant to mention that the injection process
analysis made in the previous sections of this work has been
focused mainly on two aspects: (1) the fraction of light absorbed
through FranckꢀCondon (FC) transitions that is available for
charge injection (eq 4) and (2) the fraction of excitation energy
that is ultimately localized on the ligand that is bound to the
semiconductor surface by means of the anchoring carboxy group
(eq 5). However, it is important to consider that from an
experimental point of view, although it is possible that hot
vibrational (essentially FC) states contribute to charge injection,
this process occurs mainly from the thermally equilibrated (thexi)
It is noteworthy that the emission energy of 2 is close to
the value reported for a similar complex containing only one
dimethylamino styryl substituent on the bpy fragment, instead of
3
1
two as in LH. Although in this case, dual emission was reported;
the dimethylamino styryl ligand was found to emit at 1.82 eV.
This value suggests that emission of 2 arises mainly from the
chromophoric ligand. It should also be mentioned that the
presence of COOMe groups as substituents on bipyridine, as
in 4 and 6, seems to be detrimental for the emission since in both
complexes no emission is observed. This can be due either to the
5
921
dx.doi.org/10.1021/ic1020862 |Inorg. Chem. 2011, 50, 5910–5924

Inorganic Chemistry
ARTICLE
Table 7. Distribution Function u (g ) for the HSOMO
in the semiconductor
I ðbpyÞ ¼ EðbpyÞj ðbpyÞ
T
β
Orbital, and Orbital Energetic for the Triplet State of Com-
a
plexes 1ꢀ6
ð8Þ
T
T
j
T
E(bpy) is obtained from eq 5, and values for j (bpy) are
T
b
c
obtained through the last column of Table 7. It can be seen in this
table that, as expected, complexes possessing two bpy ligands
complex
E
EM
E
T
I
T
(bpy)
L
Ru
bpy
3þ2
Ru(bpy)
2.07 2.30
1.93 1.57
1.84 1.06
1.97 1.76
0.88
100 (100)
(4, 3) exhibit higher values for j (bpy) than those with only
T
1
2
3
4
5
6
77 (95) 2 (5) 21 (0)
one (5, 6). Moreover, electron-withdrawing substituents, such as
63 (0)
36 (3)
24 (1)
56 (2)
45 (0)
0 (3) 37 (97)
0 (2) 64 (95)
COOMe, increase the values of j (bpy), as can be seen by
T
comparing 4 and 3 with 1 and 2.
0.012
0.026
0.004
0.011
Results obtained for eq 8 are displayed in the fourth column of
Table 7. It can be seen that complex 4 shows the highest value for
0 (2) 76 (97)
1.87 1.54
1.04
1 (4) 43 (94)
the bpy fragment in I followed by 3. This result coincides with
T
0 (6) 55 (94)
the tendency previously found for the FranckꢀCondon singlet in
a
L = chromophoric ligand, LH or LNMe
dicarboxybipyridine. Values for the singlet state through LUMO com-
position are displayed in parentheses for comparison purposes. All
2
, Bpy = bipyridine or
Figure 9.
b
c
energies in eV. E = E(T ) ꢀ E(T ). I (bpy) = available energy to
T
0
1
T
’ REMARKS AND CONCLUSIONS
1
be injected from the bpy fragment on the T state (eq 7).
As discussed in the text, an understanding of the spectroscopic
and electrochemical properties of the complexes could be made
with the help of theoretical calculations. This was achieved
mainly by comparing the relative energy and composition of
the MOs involved and the nature of the electronic transitions
associated to each absorption and emission band for the different
complexes.
As mentioned in the Introduction and in the cited literature
therein, ruthenium complexes bearing chromophoric ligands
should be a good choice as solar cell dyes due to the fact that
these ligands considerably increase the molar extinction coeffi-
cient values, which are directly related to the light absorbance or
light-harvesting efficiency (LHE). Taking into account the fact
that the major aim of this article is to elucidate if the absorbed
photons by this variety of dyes with chromophoric ligands are
potentially useful in the generation of electric current, the results
here reported allow one to conclude that not necessarily all the
transitions involved in the absorption bands will end in an
effective one-electron injection process to the semiconductor
conduction band. Therefore, an increase in LHE not necessarily
should imply an increase in IPCE. This general conclusion is
based on the following facts:
excited state. For ruthenium polypyridinic-type complexes, the
triplet state is a good candidate for the mentioned thexi state.
Therefore, optimization of the triplet state for the series of
compounds under study was performed. Table 7 displays
calculated triplet energies, indicated as E . The values were
T
obtained from the difference between the total energy of the
optimized triplet state and the total energy of the ground singlet
state, calculated in the triplet state optimized geometry. Avail-
þ2
able data for Ru(bpy)3 was included for comparison reasons.
A linear correlation coefficient of 0.95 was found when compar-
ing experimental emission data, EEM, with theoretical E values.
T
In accordance with this, the trend of the experimental emission
energy is correctly reproduced by the calculated values: 3 > 1 >
5
> 2. It is also predicted that complex 2 shows a lower emis-
sion energy along the series under study as well as that
NMe substitution shifts the emission band to lower energy
2
regions.
From another point of view, the composition of the highest
spin occupied molecular orbital (HSOMO) can be analyzed in
terms of molecular fragments, Table 7. Through a first examina-
tion, it can be seen that the triplet states are mostly characterized
by an increase in the participation of the chromophoric ligand
when compared to its contribution in the corresponding singlet
state (in parentheses in Table 7). More important, the chromo-
phoric ligand becomes the predominant fragment in complexes 2
and 5, where no contribution of this fragment was observed in
the singlet state. This effect can also be detected when two
chromophores are present, as can be observed on comparing
complexes 3 and 5, where an increase of 19% in the contribution
of this fragment is observed in the latter, while a comparison of
complexes 4 and 6 shows that this increase corresponds to 20%.
Note that on comparing 3 and 5 the nature of the triplet state is
quite different since in the former it is located on the bpy
fragment while for the latter it is found on the chromophore.
This difference can be related to the emission spectral shift from
1 As can be seen in Table 6, when a donor group, such as
NMe , is introduced in the chromophoric ligand, as in
2
complexes 4 and 6, an enhancement of the molar extinction
coefficient is observed for the lowest energy band, increas-
ing therefore its visible light-harvesting capacity. Never-
theless, according to Figure 8, a predominance of LL or
ILCT intraligand bands is observed for these complexes,
limiting their possibilities as dyes, as these types of absorp-
tions lack the directionality of MLCT bands.
2 Figure 8 gives the composition of the whole visible absorp-
tion framework for the different complexes in terms of the
basic transitions described in Figure 6. It can be seen that
complex 5 shows the highest contribution of MC absorp-
tion and therefore is very efficient in light harvesting, as
reflected by its high molar extinction coefficient. However,
complexes 3 and 4 possess the highest contribution of
MB transitions of all the complexes in the series studied,
Figure 8, making a direct injection processes more feasible
in these complexes. In fact, MB transitions involve an
electronic transition to the ligand, which would be anchored
to the semiconductor in a solar cell.
6
29 to 664 nm.
To understand the role of this triplet state with regard to
electronic injection efficiency, it is necessary to use eq 7 to
calculate I (bpy) in relation to the bpy fragment, i.e., the
T
fragment that contains the anchoring COOMe groups and
therefore has the possibility of direct injection of the electrons
5
922
dx.doi.org/10.1021/ic1020862 |Inorg. Chem. 2011, 50, 5910–5924

Inorganic Chemistry
ARTICLE
3
A comparison of 3 with 5 permits one to establish that
introduction of a second chromophoric ligand diminishes
the MB contribution of the band. The same is true when
comparing 4 and 6.
17, 154. (e) Nazeeruddin, M. K.; P ꢀe chy, P.; Renouard, T.; Zakeeruddin,
S. M.; Humphry-Baker, R.; Comte, P.; Liska, P.; Cevey, L.; Costa, E.;
Shklover, V.; Spiccia, L.; Deacon, G. B.; Bignozzi, C. A.; Gr €a tzel, M.
J. Am. Chem. Soc. 2001, 123, 1613. (f) Wang, Z.-S.; Yamaguchi, T.;
Sugihara, H.; Arakawa, H. Langmuir 2005, 21, 4272.
4
From remarks 2 and 3 it follows that although chromopho-
ric groups and donor substituents on the ligands are good
options to increase light-harvesting efficiency (LHE), their
possible contribution to an efficient IPCE must be analyzed
from the point of view of their contribution to the genera-
tion of excited states potentially useful for direct injection.
According to Figure 9 it can be expected that complexes 3
and 4 are most suitable to be employed as sensitizers on a
solar cell design. In fact, although complex 6 is the one with
the highest absorptionand while complex5 absorbs similar to
complex 4, they have a lower amount of available energy for
injection, as described above. It must also be considered that
according to Figure 8 complex 4 has the highest contribution
to LL (LH f bpy) bands, a fact that could increase the
electron density on the ligand where direct injection arises.
According to Table 7, when considering electronic injection
coming from the first triplet excited state, the calculated
amount of energy that can be injected suggests that com-
plexes 3 and 4 are the most suitable, in accordance with the
above discussion.
(2) (a) Le Bouder, T.; Viau., L.; Gu ꢀe gan, J.-P.; Maury, O.; Le Bozec,
H. Eur. J. Org. Chem. 2002, 3024. (b) Viau, L.; Malkowsky, I.; Costuas,
K.; Boulin, S.; Toupet, L.; Ishow, E.; Nakatani, K.; Maury, O.; Le Bozec,
H. Chem. Phys. Chem. 2006, 7, 644. (c) Hagfeld, A.; Gtzel, M. Chem. Rev.
1
995, 95, 49. (d) Wang, P.; Klein, C.; Humphry-Baker, R.; Zakeeruddin,
S. M.; Gr €a tzel, M. J. Am. Chem. Soc. 2005, 127, 808.
(3) Meyer, G. J. J. Chem. Educ. 1997, 74, 652.
(4) Jiang, K.-J.; Xia, J.-B.; Masaki, N.; Shuji, N.; Yanagida, S. Inorg.
Chem. 2008, 361, 783.
5
6
(5) (a) Aranyos, V.; Hjelm, J.; Hagfeldt, A.; Grennberg, H. Dalton
Trans 2003, 1280. (b) Jang, S.-R.; Lee, C.; Choi, H.; Ko, J. J.; Lee, J.;
Vittal, R.; Kim, K.-J. Chem. Mater. 2006, 18, 5604. (c) Handa, S.;
Wietasch, H.; Thelakkat, M.; Durranta, J. R.; Haque, S. A. Chem.
Commun. 2007, 17, 1725.
(6) Renouard, T.; Le Bozec, H.; Brasselet, S.; Ledoux, I.; Zyss, J.
Chem. Commun. 1999, 871.
(7) Le Bouder, T.; Maury, O.; Bondon, O.; Costuas, K.; Amouyal, E.;
Ledoux, I.; Zyss, J.; Le Bozec, H. J. Am. Chem. Soc. 2003, 125, 12284.
(8) (a) Nazeeruddin, M. K.; Wang, Q.; Cevey, L.; Aranyos, V.; Liska,
P.; Figgemeier, E.; Klein, C.; Hirata, N.; Koops, S.; Haque, S. A.; Durrant,
J. R.; Hagfeldt, A.; Lever, A. B. P.; Gr €a tzel, M. Inorg. Chem. 2006, 45, 787.
(
b) Angelis, F. D.; Fantacci, S.; Selloni, A.; Gr €a tzel, M.; Nazeeruddin,
M. K. Nano Lett. 2007, 7, 3189.
9) Snaith, H. J.; Karthikeyan, C. S.; Petrozza., A.; Teuscher, J.;
It must be finally pointed out that the previous analysis was
centered on direct electron injection. Nevertheless, in a cell,
through-space electronic injection could also be observed; if
this is the case, the contribution of MC absorptions to the IPCE
would become more relevant.
(
33
Moser, J. E.; Nazeeruddin, M. K.; Thelakkat, M.; Gr €a tzel, M. J. Phys.
Chem. C 2008, 112, 7562.
(10) (a) Wang, P.; Zakeeruddin, S. M.; Moser, J. E.; Humphry-
Baker, R.; Comte, P.; Aranyos, V.; Hagfeldt, A.; Nazeerunddin, M. K.;
Gr €a tzel, M. Adv. Mater. 2004, 16, 1806. (b) Wang., P.; Zakeeruddin,
S. M.; Moser, J. E.; Nazeeruddin, M.; Sekiguchi, T.; Gr €a tzel, M. Nat.
Mater. 2003, 2, 402.
’
ASSOCIATED CONTENT
S
Supporting Information. Comparison of calculated and
b
(
11) Beer, P. D.; Kocian., O.; Mortimer, R. J.; Ridgway, C. J. Chem.
Soc., Dalton Trans. 1993, 2629.
12) Cook, M. J.; Lewis, A. P.; McAuliffe, G. S. G.; Skarda, V.;
experimental UV� vis spectra for complex 3 using different
exchange-correlation functions, Figure S1, frontier orbital energy
and population, and percent of contribution of each basic
excitation to the electronic transition for complexes 1� 6, Tables
S1� S6, and Cartesian coordinates for the optimized ground and
triplet states for compounds 1� 6, Tables S7� S12 and S13� S18,
respectively. This material is available free of charge via the
Internet at http://pubs.acs.org.
(
Thomson, A. J.; Glasper, J. L.; Robbins, D. J. J. Chem. Soc., Perkin Trans II
1984, 1293.
(13) (a) Karthikeyan, C. S.; Peter, K.; Wietasch, H.; Thelakkat, M.
Sol. Energy Mater. Sol. Cells 2007, 91, 432. (b) Perez, L. C.; Kador, L.;
Peng, B.; Thelakkat, M. J. Phys. Chem. B 2005, 109, 5783.
(14) Donnici, C. L.; M ꢀa ximo, F. D. H.; Cruz, M.; Teixeira, L. L.; dos
Reis, G.; Santos, C. E.; Ferreira de Oliveira, I. M.; Carvalho, S.; Paniago,
E. B. J. Braz. Chem. Soc. 1998, 9, 455.
’
AUTHOR INFORMATION
(15) Zhou, M.; Robertson, G. P.; Roovers, J. Inorg. Chem. 2005,
4
4, 8317.
Corresponding Author
(16) Sullivan, B. P.; Salmon, D. J.; Meyer, T. J. Inorg. Chem. 1978,
*E-mail: bloeb@puc.cl.
17, 3334.
(17) Velde, G.; Bickelhaupt, F. M.; Van Gisbergen, S. J. A.; Fonseca
Guerra, C.; Baerends, E. J.; Snijders, J. G.; Ziegler, T. J. Comput. Chem.
’
ACKNOWLEDGMENT
2
002, 22, 931.
FONDECYT Grant 1070799 is gratefully acknowledged. We
would also like to acknowledge Dr. Ramiro Arratia-Perez from
UNAB for his kind support in regard to calculation facilities.
(18) Perdew, J. P.; Wang, Y. Phys. Rev. B 1992, 45, 13244.
(19) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200.
(20) Becke, A. D. Phys. Rev. B 1986, 34, 7406.
(
21) Van Lenthe, E.; Baerends, E. J. J. Comput. Chem. 2003, 24,
142.
22) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996,
77, 3865.
(23) (a) Handy, N. C.; Cohen, A. J. Mol. Phys. 2001, 99, 403. (b) Lee,
1
’
REFERENCES
(
(1) (a) Altobello, S.; Argazzi, R.; Caramori, S.; Contado, C.; Da Fr ꢀe ,
S.; Rubino, P.; Chon ꢀe , C.; Larramona, G.; Bignozzi, C. A. J. Am. Chem.
Soc. 2005, 127, 15342. (b) Nazeeruddin, M. K.; Kay, A.; Rodicio, I.;
Humphry-Baker, R.; M €u ller, E.; Liska, P.; Vlachopoulos, N.; Gr €a tzel, M.
J. Am. Chem. Soc. 1993, 115, 6382. (c) Kelly, C. A.; Farzad, F.;
Thompson, D. W.; Stipkala, J. M.; Meyer, G. J. Langmuir 1999,
C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
(24) Van Leeuwen, R.; Baerends, E. J. Phys. Rev. A 1994, 49, 2421.
(25) (a) Hirshfeld, F. L. Theor. Chim. Acta 1977, 44, 129. (b) Wiberg,
K. B.; Rablen, P. R. J. Comput. Chem. 1993, 44, 1504.
(26) Juris, A.; Campagna, S.; Bidd, I; Lehn, I.-B.; Ziessel, R. Inorg.
Chem. 1988, 27, 4007.
1
5, 7047. (d) Kuang, D.; Klein, C.; Ito, S.; Moser, J.-E.; Humphry-
Baker, R.; Zakeeruddin, S. M.; Gr €a tzel, M. Adv. Funct. Mater. 2007,
5
923
dx.doi.org/10.1021/ic1020862 |Inorg. Chem. 2011, 50, 5910–5924

Inorganic Chemistry
ARTICLE
(27) Leidner, C. R.; Sullivan, B. P.; Reed, R. A; White, B. A.;
Crimmins, M. T.; Murray, R. W.; Meyer, T. J. Inorg. Chem. 1987, 26, 882.
(
(
(
28) Kohn, W.; Sham, L. J. Phys Rev . 1965, 140, A1133.
29) Almbladth, C.; Von Barth, U. Phys. Rev. B 1985, 31, 3231.
30) (a) Barrera, M.; Zuloaga, F. Int. J. Quantum Chem. 2006,
1
06, 2044. (b) Barrera, M.; Zuloaga, F. J. Chil. Chem. Soc. 2003, 4, 27.
c) Goycolea, C.; Barrera, M.; Zuloaga, F. Int. J. Quantum Chem. 1989,
4, 123.
31) Song, L.-G.; Feng, J.; Wang, X.-S.; Yu, J.-H.; Hou, Y.-J.; Xie,
P.-H.; Zhang, B.-W.; Xiang, J.-F.; Ai, X.-C.; Zhang, J.-P. Inorg. Chem.
003, 42, 3393.
32) Han, X.; Wu, L.-Z.; Si, G.; Pan, J.; Yang, Q.-Z.; Zhang, L.-P.;
Tung, C.-H. Chem.—Eur. J. 2007, 13, 1231.
33) Gajardo, F.; Leiva, A. M.; Loeb, B.; Delgadillo, A.; Stromberg,
J. R.; Meyer, G. J. Inorg. Chem. Acta 2008, 361, 613.
(
5
(
2
(
(
’
NOTE ADDED AFTER ASAP PUBLICATION
This paper was published on the Web on May 31, 2011. Due to
a production error, Figure 2 is incorrect. The corrected version
was reposted on June 6, 2011.
5
924
dx.doi.org/10.1021/ic1020862 |Inorg. Chem. 2011, 50, 5910–5924